Saturday 30 November 2002

William Broad & Nicholas Wade: Betrayers of the Truth: Fraud and Deceit in Science (1982)

Edition: Oxford University Press, 1985 (Buy from Amazon)
Review number: 1134

Scientists have generally portrayed themselves as they have perceived themselves: as objective searchers after truth. This is itself not entirely the way things are even at the best of times, and on reflection it could hardly be the case. Scientists are human too; the ideas on which they work are the products of human minds; and total objectivity can only be an ideal to aspire to.

It is the dichotomy between the ideal and the reality which is the subject of this book - the whole spectrum of deceit in science, ranging from outright fraud to massaging of results to bolster a conclusion to subconscious self-deception. Topics touched on also include the relationships between the different people involved in research - such as senior and junior researchers, or co-authors of papers. The authors set out to determine what it is about the (then) current scientific culture which encourages fraud and its cover up, and they are particularly interested in the ways that process supposed to act as checks and balances (peer review of grant applications, refereeing of papers, and replication of results) have been corrupted. Broad and Wade seem to imply that a major source of problems - which are clearly wider than can be seen by examining only those scandals which have become public knowledge - are consequences of the professionalism of science (as it has this century become more of a career and less the province of the dedicated amateur) and the massive increase in the size of the scientific culture (they estimate that 90% of all scientists who ever worked were active at the time of publication, and that many papers and journals are virtually unread, making it easier to get away with plagiarism). This is however balanced by accounts of (relatively minor) fraud by some of the greatest names in science in previous centuries - men like Galileo, Newton and Mendel reporting experimental results unbelievably close to the predictions of their theories. The verdict of the historians of science seems to be that this is OK, provided that the theory turns out to be correct.

I'm not sure that the selection of the professionalisation of science as a major cause of fraud is entirely correct. One of the other interesting things that comes out of reading Betrayers of the Truth is that almost all the recent examples discussed come from the biological sciences, particularly medicine. This is something which the authors put down to the higher mathematics content in physics and chemistry. The twentieth century expansion of science as a whole is disproportionately centred on biology, and in medical research in particular are combined high pressure to produce results and massive rewards (both in money and status) potentially available, and considerable difficulty in designing, carrying out and correctly interpreting experiments. This is something which seems to me to be a basic part of the reason behind modern fraud, and it makes it especially tragic when flawed experiments can be used as the basis for new treatments.

It is now twenty years since the publication of Betrayers of the Truth; have things changed? I can't see that deliberate fraud and self deception will have gone away. One of the most famous episodes in science in the last few years falls pretty definitely into the latter category, for example - the story of cold fusion. That shares many features with cases described here (especially that of N-rays to which it has frequently been compared). One particular common feature is the attitude of the authorities involved, with the attempts of the university to play things up to attract grant money and the use of rhetoric rather than logic to argue the merits of the case. These are aspects that one would expect to have changed, as high profile cases of error would argue caution, but this does not seem to have happened.

Many scientists manage to go through their entire careers without coming across a case of fraud, though I suspect that most would harbour suspicion that some massaging of results has gone on at some point. One of the major surprises to me in this book is the consistent attitude of senior scientists that fraud hardly ever happens; that it is only the sick of mind who attempt it; and that accusations of fraud are best covered up. (This last is a regrettable consequence of the generally positive fact that scientists see themselves as a community.) There was an article in Physics Today (Investigation Finds that One Lucent Physicist Engaged in Scientific Misconduct) in November, which showed that lessons still hadn't been learnt - despite Broad and Wade bringing this up repeatedly twenty years ago, the investigating committee still said that the responsibilities of co-authors to check for fraud were not clear.

One of the most interesting points that Broad and Wade make is that attempting to understand how science works by discussing what ideal science would be like, as philosophers and scientists tend to do, is likely to lead to distortion of the truth and, in particular, to the reluctance to accept the existence of fraud that seems to still be rife. They would argue that the pathology of the scientific culture should be considered as well as its ideals; knowledge of how things can go wrong can help bring understanding of the healthy system. There is something in this, but it can be equally misleading to go too far the other way, basing understanding of the healthy on examination of the sick; this was a problem with the early development of psychiatry.

Betrayers of the Truth is a thought provoking and frequently shocking read, clearly written, in a journalistic style admirably appropriate to the topic.

Thursday 28 November 2002

Iris Murdoch: The Italian Girl (1964)

Edition: Chatto & Windus, 1964 (Buy from Amazon)
Review number: 1133

All of the Murdoch novels I have read have melodramatic elements to their plots - and, indeed, much of their effectiveness comes from the conjunction of her placid, poetic prose style with the events described - but The Italian Girl is the most fantastical of them all.

The narrator, Edmund - and it's a common trick for Murdoch, though rare in other writers, to have a narrator of the opposite sex - returns to the home where he grew up, following the death of his mother. The family, dominated by the suffocatingly unmaternal Lydia, was what would today be described as dysfunctional, and Edmund has not returned since he was initially able to escape. In doing so, he abandoned his brother Otto and Otto's wife and daughter to their fates, along with the last of the series of Italian girls who had acted as the brothers' nannies. The melodrama basically comes from the relationships between Otto's family and his apprentice, an amoral young man who lives with his sister in the grounds of the house.

A large part of the description of the relationships in the novel seems to act as a symbol for something else. This symbolic aspect of the novel is not attached to some external object as it is in The Bell or The Sea, The Sea, which makes it more difficult to guess at what is signified. The house is some kind of prison, the apprentice and his sister some kind of infection, or possibly a dangerous freedom, and the point of the whole is something to do with morality, but more than that I cannot say.

The Italian Girl is not one of Murdoch's greatest novels, but it catches the reader up like a whirlwind once it gets going. Her best books are fascinating; this one is involving but not that thought provoking.

Friday 22 November 2002

Isaac Asimov: Foundation's Edge (1982)

Edition: Granada, 1984 (Buy from Amazon)
Review number: 1132

Foundation's Edge was the first science fiction novel Asimov had written for a decade, most of which he had spent concentrating on non-fiction. It had a mixed reception, composed on the one hand of the desire to extend a welcome to an old friend long absent (which brought it the Hugo award for 1983), and on the other of the feeling that it was a sequel which failed to live up to the classic Foundation trilogy which it follows.

Reading Foundation's Edge now, after the fuss has died down, and doing so just after revisiting the original trilogy, I am more inclined to the former view, changing my mind after first acquaintance twenty years ago. This is because the earlier books have now come to seem dated, like a lot of Asimov's early fiction.

Foundation's Edge is set about halfway between the creation of the Foundation and its predicted establishment of a new empire a thousand years after the fall of the old one. (This little piece of chronology makes it clear just how large a scope Asimov had left himself for a sequel.) The Mule is long defeated, and the Second Foundation has returned to obscurity, convincing the Foundation that it too has been destroyed. Seldon's plan is back on course - and this eventually provokes suspicion in both Foundations; after so great a disruption as the reign of the Mule, how can centuries old predictions suddenly become minutely valid once again? This prompts both Foundations to begin searching for whoever or whatever has caused this, and this search is what Foundation's Edge is about.

Asimov's science fiction revolves around two great ideas: the laws of robotics and the science of psychohistory. In practice, the main interest of the plots he devises using these ideas is the ways he finds to circumvent their limitations - almost all the robot stories are about attempts to bend or break the laws of robotics, and the Foundation stories are about applying the laws of psychohistory to small numbers of individuals (because novels need to have personalities in them), something explicitly forbidden by the statistics on which the rules are supposedly based. (Alternatively, he allows an individual like the Mule to overturn the predictions, using precisely the justification that it is impossible to apply psychohistory to the actions of individuals.) One of the biggest problems that the original trilogy has, it now sdeems to me, is that the tension this produces is not fully integrated into the plot; in Foundation's Edge, it is handled much more expertly, as befits a writer with thirty years' more experience.

Asimov's characterisation is generally pretty perfunctory (the most obvious exceptions being Elijah Baley and the members of the Black Widowers), but here it is rather better than usual, the personalities of those involved playing an important part in the way that the plot is resolved.

The greater maturity of the writing should ensure that Foundation's Edge dates more slowly than its predecessors (though they, of course, first appeared over four decades ago). However, it is still the idea behind the series as a whole which is of central interest, much more so than the merits of the individual novels. The sweep of galactic history and the interest of psychohistory will probably mean that the original trilogy continues to be read for some time yet.

Saturday 16 November 2002

Lois McMaster Bujold: The Curse of Chalion (2001)

Edition: Voyager, 2002
Review number: 1131

With The Curse of Chalion, Bujold takes a break not just from her long-running and acclaimed Vorkosigan series, but also from futuristic science fiction; it is a fantasy novel with a background more or less traditional for the genre.

The novel begins with Chalion aristocrat Cazilar returning to the land of his birth; betrayed into slavery by a corrupt general, he has recently escaped from the galleys of Roknari, and now his only ambition is to see if he can persuade an old patron to take him on in a menial position in her household. Instead, he is made tutor-secretary to her granddaughter, and becomes involed in the complex politics of the Chalionian court - where he has to meet the enemy who betrayed him.

The Vorkosigan series has already shown that Bujold excels at the creation of political thrillers, and she uses the same skills here, with different characters and background. She is just as interesting and convincing writing about magic as she is discussing cloning. This particularly story is very self-contained, not only unusual in a genre which runs to trilogies and series, but also in contrast to Bujold's other work. A sequel seems unlikely, but maybe we can hope for more stories set in this world.

A lot of science fiction and fantasy contains depictions of political intrigue, an inevitable part of writing about fictional societies. The problem tends to be that it is easy enough to invent and describe institutions, but rather more difficult to portray plotting and machination convincingly. This actual human interaction behind the scenes is really what politics is about. (This is partly because real world equivalents are usually kept as secret as possible, so that it requires considerable insight to recreate convincingly, particularly in an alien setting.) One fairly common mistake that writers make is to continually describe a character as some kind of political genius, while making them achieve coups which are comparatively simple minded; this may flatter the reader but is not really very impressive when they reflect on what has been described. (Frank Herbert, who is particularly associated with the introduction of politics into science fiction, is sometimes guilty of this error.) Bujold's politicians are much more believable and her strengths in characterisation mean that she doesn't have any need to resort to this spurious trick.

Her novels also always have an exciting plot to draw the reader in. In other words, her novels are top rank political thrillers, and The Curse of Chalion is no exception. Once again, Bujold has produced a novel which tempted me into staying up into the night to finish it.

Tuesday 5 November 2002

Paul J. McAuley: Child of the River (1997)

Edition: Gollancz, 1997
Review number: 1130

The strange world of Confluence is the setting for McAuley's trilogy of the same name; Child of the River is the first volume. It was created by the Preservers, who populated it with diverse creatures derived from Earth's animals raised to human intelligence, and who then disappeared. Confluence is now considerably decayed, full of machines whose purpose is not understood, whose workings follow the Arthur C. Clarke dictum and are indistinguishable from magic; while the Preservers themselves have been mythologised into a pantheon of gods supposedly watching over the fate of the inhabitants.

The central character, the boy Yama, appears set apart from the beginning; Moses-like, he is found as a baby floating down the great River, not in a reed basket but accompanying the coffin of a dead young woman. He is brought up as the adopted son of a village official, but becomes the centre of intrigue when others start to investigate his "bloodline" - he is not one of the common types found throughout Confluence, and so his heritage is unknown. He finds that he experiences a strange communion with the machines around him which marks him out, and he begins to travel to find out who he is and why he is there.

The plot clearly picks up many ideas from Joseph Campbell's The Hero With a Thousand Faces, but it is the setting and style which are of greater interest. The latter is distincly reminiscent of Gene Wolfe, a writer greatly admired but not, I think, frequently imitated. There are two sequences of novels which Child of the River calls to mind: The Book of the New Sun and, even more so, The Book of the Long Sun (which is more or less contemporary with this trilogy). This similarity is partly because of connections between the backgrounds of the series, but it is mainly because of the way that the background is expressed in the story. (The down side of this, more obviously so in McAuley than Wolfe, is that exposition of the setting and delight in detail get in the way of plot development.) I would not say that McAuley is as good a writer as Wolfe, not that Child of the River is his best work; it is a little too derivative for that. It is, though, an interesting and worthwhile read.